MATLAB: How to add variables to the Plot Fcn

addfminconfminsearchMATLABoptimsetoptionsplotfcnplotfcnsvariable

I created a custom Plot Function for 'fminsearch':

opts = optimset('Display','iter','TolFun',1e-5,'MaxIter',100,'PlotFcns',@outf);
[v1, fval] = fminsearch(funtol,v1,opts);
function stop = outf(x, optimValues, state)
...<Plot some stuff>...
end

How can I add some variables to the Plot Function? For example:

function stop = outf(x, optimValues, state, var1, var2)

I don't know how to call it then in 'optimset' properly, because, it seems, there are no additional varibles allowed:

opts = optimset('PlotFcns',@outf<Don't know what to write here>);

Best Answer

One solution is to write an anonymous function and define the extra values prior to creating the anonymous function. Those extra values will then be stored in that function until it's deleted or overwritten. Example:

% Assuming your outf function is within the function or script
var1 = 0; 
var2 = 1; 
outfAnon = @(x) outf(x, optimValues, state, var1, var1);
opts = optimset('PlotFcns',outfAnon, ...)

For more information on this method or to see other methods including the use of nested functions and global vars (avoid), see this : https://www.mathworks.com/help/optim/ug/passing-extra-parameters.html

[EDIT] Removed a mistake where I included ' @' symbol when entering outfAnon in the opts line.

Related Solutions

MATLAB: How to conveniently plot iteration progress during a 2D optimization

First, you should create your own function for plotting a 3D contour during the optimization process:

function stop = custom_plot(x, optimValues, state, <list of your custom parameters>)
    
    % Standard part
    stop = false;
    if state ~= 'iter'
        return
    end
    
    % This will clear the previously plotted contour if needed
    if optimValues.iteration > 0
        cla;
    end
    
    % Here insert your code as if you plotted a graph normally, like 'plot' or 'surf', etc.
    % Don't forget that 'x' is your vector of two variables
    % Also consider looking into 'optimValues' structure, optimValues.fval contains your current function value
    % Feel free to use any custom parameters from <list of your custom parameters>
    
    <...>
        
    % Consider trying surf(x(1), x(2), optimValues.fval);
    
end

Second, in your main part of code you should call the 'custom_plot' function like this:

outfanon = @(x, optimValues, state) custom_plot(x, optimValues, state, <list of your custom parameters>);
opts = optimset('Display','iter','TolFun',1e-5,'TolX',1e-5,'MaxIter',100,'PlotFcns',outfanon);
[x1, fval] = fminsearch(f, x0, opts);

MATLAB: Fmincon check Hessian (in DerivativeCheck)

I don't think you can, but an indirect way would be to run like below with MaxIter=1 and have fmincon return the Hessian with the 'Hessian' option turned both on and off. Then you can check the Hessian error manually.

I agree it is counter-intuitive, but I think it might be deliberate that the DerivativeCheck does not test the Hessian. Often, minimization algorithms will work with fairly crude approximations to the Hessian, so maybe it is expected that your user-supplied Hessian will not always be exact. Or, maybe it is with large dimensional problems in mind, where finite difference derived Hessians are too expensive to compute, even for testing purposes, but you still might want to use DerivativeCheck to verify the gradient.

Q=[2 1;1,2];
 Q=Q+Q.';
   opts=optimoptions(@fmincon,'DerivativeCheck','on',...
                              'GradObj','on','Hessian','off','MaxIter',1);
  [x,fval,exitflag,output,lambda,grad,H_off]=...
      fmincon(@(x)objfun(x,Q),[1,2],[],[],[],[],[-1,-1],[1,1],[],opts)
   opts=optimoptions(@fmincon,'DerivativeCheck','on',...
                              'GradObj','on','Hessian','on','MaxIter',1);
[x,fval,exitflag,output,lambda,grad,H_on]=...
    fmincon(@(x)objfun(x,Q),[1,2],[],[],[],[],[-1,-1],[1,1],[],opts)
HessianError=H_on-H_off
keyboard
function [f,g,H]=objfun(x,Q)
   x=x(:);
   f=x.'*Q*x/2;
   g=Q*x;
   H=eye(2);

Best Answer

Related Solutions

MATLAB: How to conveniently plot iteration progress during a 2D optimization

MATLAB: Fmincon check Hessian (in DerivativeCheck)

Related Question